Risk factors for relapse and nomogram for relapse probability prediction in patients with minor ischemic stroke

Multimodal ischemic stroke recurrence prediction model based on the capsule neural network and support vector machine

Ischemic stroke (IS) has a high recurrence rate. Machine learning (ML) models have been developed based on single-modal biochemical tests, and imaging data have been used to predict stroke recurrence. However, the prediction accuracy of these models is not sufficiently high. Therefore, this study aimed to collect biochemical detection and magnetic resonance imaging (MRI) data to establish a dataset and propose a high-performance heterogeneous multimodal IS recurrence prediction model based on deep learning. This is a retrospective cohort study. Data were retrospectively collected from 634 IS patients in Zhuhai, China, a 12-month follow-up was conducted to determine stroke recurrence. We propose the ischemic stroke multi-group learning (ISGL) model, an integrated model for predicting the recurrence risk of multimodal IS in patients, based on a capsule neural network and a linear support vector machine (SVM). Two capsule neural network prediction models based on T1 and T2 signals in the MRI data and a SVM prediction model based on biochemical test data were established. Finally, a vote was conducted on the final judgment of the integrated model. The ISGL model was compared with 6 classical ML and deep learning models: k-nearest neighbors, SVM, logistic regression, random forest, eXtreme Gradient Boosting, and visual geometry group. The results revealed that the accuracy, specificity, sensitivity and the area under the curve of the ISGL model were 95%, 96%, 94%, and 95%, respectively. Among the comparison models, the visual geometry group method exhibited the best performance, but it much lower than those of the ISGL model. Analysis of the importance of biochemical test data revealed that low-density lipoprotein, smoking, and heart disease history were the positively correlated factors, and total cholesterol, high-density lipoprotein, and diabetes were and the negatively correlated factors. This study proposes the ISGL model can be used simultaneously with MRI and biochemical data to predict IS recurrence. This combination resulted in higher rate of performance than that of the other ML models. Additionally, this study found related risk factors affected recurrence, which can be used to intervene in high-risk patients' recurrence as early as possible and promote the development of secondary prevention of stroke.

Nomogram to predict hemorrhagic transformation for acute ischemic stroke in Western China: a retrospective analysis

BACKGROUND AND PURPOSE

Hemorrhagic transformation (HT) is the most alarming complication of acute ischemic stroke. We aimed to identify risk factors for HT in Chinese patients and attempted to develop a nomogram to predict individual cases.

METHODS

A retrospective study was used to collect the demographic and clinical characteristics of ischemic stroke patients at the Second Affiliated Hospital of Chongqing Medical University (development cohort) and Chongqing Sanbo Changan Hospital (validation cohort) from October 2013 to August 2020. Univariate analysis and multivariate analysis were used to identify the risk factors of patients in the development cohort. The nomogram was generated, and internal validation was performed. We used the area under the receiver-operating characteristic curve (AUC-ROC) to assess the discrimination and used the Hosmer-Lemeshow test to calibrate the model. To further verify the predictability and accuracy of the model, we performed an external validation of the patients in the validation cohort.

RESULTS

A total of 570 patients were used to generate the nomogram. After univariate analysis and multivariate logistic regression, the remaining 7 variables (diabetes mellitus, atrial fibrillation, total cholesterol, fibrous protein, cerebral infarction area, NIHSS score and onset-to-treatment) were independent predictors of HT and used to compose the nomogram. The area under the receiver-operating characteristic curve of the model was 0.889 (95% CI, 0.841-0.938), and the calibration was good (P = 0.487 for the Hosmer-Lemeshow test). The model was validated externally with an AUC-ROC value of 0.832 (95% CI, 0.727-0.938).

CONCLUSIONS

The nomogram prediction model in this study has good predictive ability, accuracy and discrimination, which can improve the diagnostic efficiency of HT in patients with acute ischemic stroke.